AI's Surveillance Surge: 86% of Computer Vision Patents Focus on Human Imaging, Study Reveals
June 25, 2025
The research analyzed over 19,000 computer-vision papers and 23,000 related patents, revealing that a staggering 86% of these patents involve human imaging data, with only 1% focusing on non-human data.
Advancements in AI and imaging have enabled sophisticated human-surveillance capabilities, including facial and gait recognition, which proponents argue enhance public safety.
The study concludes that the computer vision field systematically engages in the extraction of human data, often neglecting ethical considerations like consent and privacy, thereby contributing to a culture of surveillance.
A new study published in Nature highlights that the majority of research papers and patents in artificial intelligence, particularly in computer vision, are predominantly focused on developing surveillance technologies.
This trend towards surveillance-related patents has intensified over the decades, with 78% of computer-vision papers leading to patents related to surveillance in the 2010s, a significant increase from 53% in the 1990s.
However, researchers express concerns that these technologies could infringe on democratic freedoms by facilitating the identification and tracking of individuals.
Past instances of data collection without consent, such as the Duke University case, have raised alarms about ethical standards in AI research.
Various stakeholders, including grassroots communities, academics, and legislators, are increasingly concerned about how computer vision technologies infringe on privacy and exacerbate societal disparities.
The podcast accompanying the study encourages further exploration of the ethical implications of AI in surveillance, emphasizing the need for transparency and active engagement from researchers.
The editorial warns that referring to humans as 'objects' in research can obscure the potential consequences of how surveillance technology is applied.
Despite some beneficial applications of computer vision in fields like healthcare, the findings underscore a troubling pipeline from research directly to surveillance technologies.
Researchers are urged to adhere to ethical standards, clearly define the human element in their studies, and engage in the broader societal debate regarding the implications of their work.
Summary based on 6 sources
Get a daily email with more AI stories
Sources

Nature • Jun 25, 2025
Is AI powering Big Brother? Surveillance research is on the rise
Nature • Jun 25, 2025
Computer-vision research is hiding its role in creating ‘Big Brother’ technologies
Nature • Jun 25, 2025
Computer-vision research powers surveillance technology
Nature • Jun 25, 2025
Is AI watching you? The hidden links between research and surveillance